Light emitting diodes (LEDs) have been developed for the next generation of visible light communication (VLC) in addition to improved LEDs for illumination. This means that communication to and from LED based infrastructures such as traffic lights, lamps, car headlights, street lights and electronic posters is now possible and would provide an alternative form of communication in RF-restricted domains, such as airplanes, military applications, subsea equipments and hospitals.
For VLC, high brightness LEDs with the ability to switch on and off quickly is required to transmit large volumes of data over large distances. However the slow spontaneous emission (SE) (e.g., in the range of 2-4 ns) from LEDs places a fundamental limit on their modulation speed to 250 MHz to 500 MHz. In order to push the speed towards the GHz or even multi-GHz range, it is necessary to increase the intrinsic SE rate. One approach is to increase the dopant concentration of the active material. By using a heavily doped active region (e.g., in the range of 1019 cm3), the carrier radiative lifetime can be reduced to 100 ps, corresponding to 1.6 GHz. However, increasing the dopant concentration also introduces non-radiative recombination centers within the active material, which severely limits the internal efficiency of the LEDs.
By modifying the electromagnetic field in the vicinity of the emitter, it is also possible to increase the SE rate. This is commonly achieved by using a cavity or photonic crystal design. Such a design can be implemented in a microcircuit but scalability is difficult for large scale production. Resonant cavity LEDs, which involve growing a set of Bragg reflectors, can help shorten the lifetime due to resonant cavity effect. However, the enhancement occurs mainly along the cavity axis and emission rates along other directions are not significantly changed. Therefore, the average spontaneous lifetime over all angles is reduced by only 10%.
Using surface plasmon (SP) coupling to increase the SE rate is another approach. This can be done by simply depositing a metal layer, usually Ag, on top of a LED device. The electron-hole pairs in the InGaN quantum wells (QW) can then recombine to form surface plasmons (SP) at the metal/GaN interface instead of being emitted as free photons. In this manner, the electron-hole recombination rate is significantly increased because it emits as a SP instead of a photon. This new path of recombination can also significantly increase the SE rate due to the much higher density of the SP states. However, the SP fringing field at GaN/Ag decreases exponentially with distance from the interface. This places a restriction on the LED device fabrication, as it means that the p-GaN layer will have to be ultrathin. Typically, the thickness of p-GaN needs to be thicker than 76 nm for InGaN LED to maintain the efficiency needed to form a p-n junction in LED.
Thus, what is needed is a high SE-rate LED with high efficiency. Furthermore, other desirable features and characteristics will become apparent from the subsequent detailed description, taken in conjunction with the accompanying drawings and this background of the disclosure.